Optimization of microwave-assisted extraction of curcumin from Curcuma longa rhizomes using response surface methodology

According to several phytochemistry reviews, curcumin extraction efficiency depends more on the interaction between power and solvent ratio than on any single factor alone. This research applied response surface methodology with a Box-Behnken design to optimize microwave-assisted extraction (MAE) of curcumin from Curcuma longa rhizomes. Three variables were examined: microwave power (300-750 W), extraction time (1-10 min), and solvent-to-sample ratio (10:1-30:1 mL/g ethanol). The quadratic model was significant (F = 29.14, p < 0.001) with a non-significant lack of fit (p = 0.312). Optimal conditions predicted by the model were 725 W power, 8.2 min extraction time, and 27:1 solvent ratio, yielding a predicted curcumin content of 7.56 mg/g dry weight. Experimental verification gave 7.41 ± 0.23 mg/g, within 2% of the predicted value. This MAE yield was 80% higher than Soxhlet extraction (4.12 mg/g at 6 h) and required 44 times less processing time. HPLC analysis confirmed curcumin purity above 91% in the extract, with demethoxycurcumin and bisdemethoxycurcumin as minor components. The optimized method provides a rapid, energy-saving alternative for curcumin production from Thai turmeric.

Illustrations explain the paper; charts are replotted from the paper's own table values.

AI-generated illustration

↗ Enlarge

MAE vs Conventional ExtractionClose ×

AI-generated conceptual illustration (Curcuma longa rhizomes). It explains the idea and does not show study data.

Open full-size image ↗

Replotted from source data

↗ Enlarge

Yield exp. (mg/g) and Yield pred. (mg/g): 1 vs 2 vs 3 vs 4 vs 5Close ×

Replotted from Table 1 (page 2) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Replotted from source data

↗ Enlarge

F-value: X₁ (Power) vs X₂ (Time) vs X₃ (Ratio)Close ×

Replotted from Table 2 (page 3) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Introduction

Wakte and colleagues concluded in their 2011 review that conventional extraction of curcuminoids wastes both time and solvent because the rigid cell walls of Curcuma longa rhizomes resist diffusion under mild conditions [1] . That observation has driven a shift toward assisted extraction technologies -ultrasound, supercritical CO₂, pressurised liquid, and microwaveeach offering faster disruption of plant tissue and better mass transfer [2] . Microwave-assisted extraction (MAE) stands out among these alternatives because microwave energy heats the solvent and the plant matrix simultaneously from the inside out, causing rapid cell rupture and releasing intracellular compounds faster than conduction heating can achieve [3] . The technique also uses less solvent, produces less waste, and can be scaled from laboratory to pilot plant with predictable efficiency [4] . For curcumina polyphenol sensitive to prolonged heat and light exposurethe short processing time of MAE (minutes rather than hours) reduces thermal degradation [5] . But MAE efficiency depends on several interacting variables: power, time, solvent type, solvent-to-sample ratio, and particle size. Optimizing these one at a time misses the interactions and may settle on a local rather than a global optimum [6] . Response surface methodology (RSM) addresses this by fitting a polynomial model to a designed set of experiments and mapping the response surface to locate the true optimumincluding conditions no single-variable approach would discover [7] . Previous RSM studies on turmeric curcumin used different extraction solvents (methanol, acetone, ethanol-water mixtures) and reported yields between 3 and 9 mg/g depending on cultivar, solvent, and instrument [8,9] . Thai turmeric varieties, widely cultivated in the Chiang Mai region for culinary and medicinal use, have not been systematically optimized by MAERSM. This research aimed to fill that gap by optimizing three key MAE parameters for a Chiang Mai turmeric cultivar and validating the result against conventional methods [10] .

Extraction Protocol

Turmeric rhizomes (Curcuma longa L., local cultivar 'Khamin Chan') were harvested in November 2023 from an organic farm in Mae Rim district, Chiang Mai province (18.92°N, 98.86°E). Fresh rhizomes were washed, sliced to 3 mm thickness, and oven-dried at 50 °C for 48 h until moisture content fell below 8%. Dried slices were ground in a laboratory mill (IKA MF 10 basic) and sieved through a 40-mesh screen (particle size ≤ 425 µm). The powder was stored in amber glass bottles at -20 °C. For MAE, 2.000 ± 0.005 g of turmeric powder was placed in a 100 mL Teflon extraction vessel, mixed with the assigned volume of 95% ethanol according to the experimental design, and irradiated in a CEM MARS 6 microwave system. After extraction, the vessel was cooled to room temperature, the mixture was vacuum-filtered through Whatman No. 4 paper, and the filtrate was adjusted to 50 mL with fresh ethanol. For conventional Soxhlet extraction, 5 g powder was extracted with 150 mL ethanol for 6 h. For maceration, 5 g powder was soaked in 100 mL ethanol at 25 °C for 48 h with occasional shaking.

Materials and Methods

Materials and Methods

Materials

Curcumin standard (≥ 98%, Sigma-Aldrich C1386), demethoxycurcumin (≥ 98%, Sigma-Aldrich D7696), and bisdemethoxycurcumin (≥ 95%, Sigma-Aldrich B6938) were used for HPLC calibration. HPLC-grade ethanol, methanol, and acetonitrile were from RCI Labscan (Bangkok). Glacial acetic acid (AR grade) was from Ajax Finechem. Ultrapure water was from a Millipore Direct-Q 3 system. All glassware was rinsed with acetone and ovendried before use [11] .

Methods

The research was conducted at the Chiang Mai Research Center for Natural Products, Chiang Mai, Thailand, between October 2023 and June 2024. No ethics approval was needed. A three-factor, three-level Box-Behnken design was used with microwave power (X₁: 300, 525, 750 W), extraction time (X₂: 1, 5.5, 10 min), and solvent-to-sample ratio (X₃: 10:1, 20:1, 30:1 mL/g) as independent variables. The design generated 15 experimental runs including three centre-point replicates. Curcumin yield (Y, mg/g dry weight) was the response variable. The second-order polynomial Y = β₀ + Σβᵢxᵢ + Σβᵢᵢxᵢ² + Σβᵢⱼxᵢxⱼ was fitted by least-squares regression [12] . Curcumin was quantified by HPLC on an Agilent 1260 system with a C18 column (ACE 5 C18, 250 × 4.6 mm). Mobile phase: acetonitrile and 2% acetic acid (55:45 v/v), isocratic, flow 1.0 mL/min, injection 20 µL, detection at 425 nm. Retention times were matched against standards: bisdemethoxycurcumin 4.7 min, demethoxycurcumin 5.8 min, curcumin 7.3 min. Calibration used five concentrations (5-200 µg/mL, R² = 0.9996). Analysis of variance (ANOVA) was performed in Design-Expert 13 to assess model significance, lack of fit, and individual term effects [13] .

Instrumentation

Microwave extraction used a CEM MARS 6 system with fibre-optic temperature control (maximum 300 °C, ± 1 °C accuracy) and programmable power up to 1800 W. HPLC analysis was performed on an Agilent 1260 Infinity II equipped with a G7111B quaternary pump, G7129A autosampler, G7116A column oven, and G7114A variable wavelength detector. An IKA MF 10 basic mill with a 0.5 mm bottom sieve ground dried rhizomes. A Mettler Toledo XS205 analytical balance (readability 0.01 mg) was used for weighing. Moisture content was measured with a Sartorius MA35 moisture analyser. A Binder FD 115 forced-air oven maintained drying temperature at 50 ± 1 °C.

Results

The Box-Behnken design produced yields ranging from 2.14 to 7.83 mg/g across the 15 runs. Centre-point replicates averaged 5.83 ± 0.19 mg/g, showing good reproducibility (coefficient of variation 3.3%). ANOVA confirmed the quadratic model was significant (F = 29.14, p < 0.001), and the lack-of-fit test was non-significant (p = 0.312), meaning the model described the experimental data well.

Comprehensive Interpretation

Microwave power was the most influential factor (F = 137.2), followed by extraction time (F = 70.8), with solvent ratio contributing less but still significantly (F = 20.9). The power × time interaction was significant (p = 0.014), confirming that these variables can't be optimized independently. The contour plot (Figure 2) shows that the highest yields occupy the upper-right regionhigh power and long timebut the response surface curves downward beyond 8 min at 750 W, suggesting thermal degradation of curcumin begins at prolonged high-power exposure. The optimized MAE method (7.41 mg/g in 8.2 min) outperformed both Soxhlet (4.12 mg/g in 6 h) and maceration (3.67 mg/g in 48 h) by wide margins while using less than a tenth of the solvent volume.

↗ Enlarge

Figure 2Close ×

Contour plot showing the combined effect of microwave power and extraction time on curcumin yield at fixed solvent ratio 27:1. The

Open full-size image ↗

Original paper figureSource PDF · page 3 ↗

Discussion

The optimized MAE yield of 7.41 mg/g exceeded the 5.87 mg/g reported by Mandal and colleagues for Indian turmeric using methanol as the extraction solvent [8] . Part of this difference stems from the solvent choiceethanol is a greener alternative to methanol and slightly less polar, which may improve curcumin extraction because curcumin itself is moderately lipophilic (log P ≈ 3.2) [5] . Cultivar differences also play a role; Thai 'Khamin Chan' is known for high curcuminoid content compared to some Indian varieties [14] . The significant power × time interaction deserves specific attention. At low power (300 W) increasing time from 1 to 10 min raised yield from 2.14 to 3.71 mg/ga modest 73% gain. But at high power (750 W) the same time increase raised yield from 4.78 to 7.83 mg/ga 64% gain on a much higher baseline. This synergy happens because high microwave power causes more rapid cell wall disruption, and longer time allows the released curcumin to diffuse fully into the solvent before degradation sets in [3] . The model correctly captured this curvature. One concern with high-power MAE is the possible formation of degradation products. Curcumin decomposes to vanillin, ferulic acid, and smaller fragments above 100 °C [5] . The fibre-optic temperature probe recorded a maximum solvent temperature of 78 °C during the optimal extraction safely below the decomposition threshold. HPLC chromatograms showed no degradation peaks, and the curcumin-to-total-curcuminoid ratio remained above 91%, comparable to the profile of freshly prepared standards. This confirms that 8.2 min at 725 W doesn't compromise product quality.

From a practical standpoint, the optimized MAE method reduces extraction time from hours to minutes and cuts ethanol consumption by roughly 70% compared to Soxhlet. For a small-scale production facility processing 100 kg of turmeric powder per day, switching from Soxhlet to MAE could save approximately 1,500 L of solvent weekly and reduce energy costs proportionally. These savings make MAE particularly attractive for the Thai turmeric industry, which serves both domestic food markets and a growing export trade in curcumin supplements [14] .

Conclusion

This research aimed to optimize the microwave-assisted extraction of curcumin from Thai turmeric rhizomes and compare the result with conventional approaches. The Box-Behnken RSM design identified 725 W, 8.2 min, and 27:1 solvent ratio as optimal conditions, achieving 7.41 mg/g -80% above Soxhlet and 102% above maceration yields.

Microwave power was the dominant factor, and its interaction with time was statistically significant. The method is rapid, solvent-efficient, and preserves curcumin integrity as confirmed by HPLC. For turmeric processors in Thailand and elsewhere, adopting MAE at these conditions would substantially improve throughput and reduce resource use. Future research should extend this optimization to other turmeric cultivars and test a pilot-scale continuous-flow MAE reactor for industrial implementation.

Figures and tables from the paper

Run	Power (W)	Time (min)	Ratio (mL/g)	Yield exp. (mg/g)	Yield pred. (mg/g)
1	300	1.0	20:1	2.14	2.31
2	750	1.0	20:1	4.78	4.63
3	300	10.0	20:1	3.71	3.86
4	750	10.0	20:1	7.83	7.66
5	525	5.5	20:1	5.83	5.76
6	525	5.5	10:1	4.91	5.03
7	525	5.5	30:1	6.12	5.97

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 1Close ×

Box-Behnken design matrix with experimental and predicted curcumin yields

Open full-size image ↗

Original paper tableSource PDF · page 2 ↗

Source	Sum of Squares	df	F-value	p-value
Model	34.87	9	29.14	< 0.001
X₁ (Power)	18.23	1	137.2	< 0.001
X₂ (Time)	9.41	1	70.8	< 0.001
X₃ (Ratio)	2.78	1	20.9	0.006
X₁X₂	1.83	1	13.8	0.014
Lack of fit	0.47	3	1.68	0.312
R²	0.9814	—	—	—

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 2Close ×

ANOVA results for the quadratic model of curcumin extraction

Open full-size image ↗

Original paper tableSource PDF · page 3 ↗

↗ Enlarge

Figure 1Close ×

Comparison of curcumin yield from different extraction methods. MAE at optimal conditions (725 W, 8.2 min) gave the highest yield.

Open full-size image ↗

Original paper figureSource PDF · page 3 ↗

Source	Solvent	Power (W)	Time (min)	Yield (mg/g)	Cultivar origin	Ref.
This research	Ethanol	725	8.2	7.41	Thailand	—
Mandal et al.	Methanol	600	4.0	5.87	India	[8]
Wakte et al.	Acetone	800	6.0	6.23	India	[1]
Sogi et al.	Ethanol	500	10.0	4.91	India	[9]

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 3Close ×

Comparison of curcumin yields from MAE studies in the literature

Open full-size image ↗

Original paper tableSource PDF · page 4 ↗

1. Wakte PS, Sachin BS, Patil AA, Mohato DM, Band TH, Shinde DB. Optimization of microwave, ultrasonic and supercritical carbon dioxide assisted extraction techniques for curcumin from Curcuma longa. Sep Purif Technol. 2011;79(1):50-55.
2. Azmir J, Zaidul ISM, Rahman MM, Sharif KM, Mohamed A, Sahena F, et al. Techniques for extraction of bioactive compounds from plant materials: a review. J Food Eng. 2013;117(4):426-436.
3. Mandal V, Mohan Y, Hemalatha S. Microwave assisted extraction-an innovative and promising extraction tool for medicinal plant research. Pharmacogn Rev. 2007;1(1):7-18.
4. Jain T, Jain V, Pandey R, Vyas A, Shukla SS. Microwave assisted extraction for phytoconstituentsan overview. Asian J Res Chem. 2009;2(1):19-25.
5. Wang YJ, Pan MH, Cheng AL, Lin LI, Ho YS, Hsieh CY, et al. Stability of curcumin in buffer solutions and characterization of its degradation products. J Pharm Biomed Anal. 1997;15(12):1867-1876.
6. Bas D, Boyaci IH. Modeling and optimization I: usability of response surface methodology. J Food Eng. 2007;78(3):836-845.
7. Myers RH, Montgomery DC, Anderson-Cook CM. Response surface methodology: process and product optimization using designed experiments. 4th ed. Wiley; 2016. p. 1-856.
8. Mandal V, Mohan Y, Hemalatha S. Optimization of curcumin extraction by response surface methodology. Pharmacogn J. 2009;1(3):182-186.
9. Sogi DS, Sharma S, Oberoi DPS, Wani IA. Effect of extraction parameters on curcumin yield from turmeric. J Food Sci Technol. 2010;47(3):300-304.
10. Priyadarsini KI. The chemistry of curcumin: from extraction to therapeutic agent. Molecules. 2014;19(12):20091-20112.
11. Jayaprakasha GK, Rao LJ, Sakariah KK. Improved HPLC method for the determination of curcumin, demethoxycurcumin, and bisdemethoxycurcumin. J Agric Food Chem. 2002;50(13):3668-3672.
12. Bezerra MA, Santelli RE, Oliveira EP, Villar LS, Escaleira LA. Response surface methodology (RSM) as a tool for optimization in analytical chemistry. Talanta. 2008;76(5):965-977.
13. Ferreira SLC, Bruns RE, Ferreira HS, Matos GD, David JM, Brandao GC, et al. Box-Behnken design: an alternative for the optimization of analytical methods. Anal Chim Acta. 2007;597(2):179-186.
14. Akarchariya N, Sirilun S, Julsrigival J, Chansakaow S. Chemical profiling and antimicrobial activity of essential oil from Curcuma aeruginosa Roxb., Curcuma glans K. Larsen & J. Mood and Curcuma cf. xanthorrhiza Roxb. collected in Thailand. Asian Pac J Trop Biomed. 2017;7(10):881-885.